Some speech recognition systems may operate in multi-user environments, where voice commands may be provided from multiple users at various volumes and from various locations within a room. In some situations it may be difficult to determine which user is currently interacting with the speech recognition system, and/or whether a speech input is being heard and understood by a computing device receiving the speech input.